中国科学院软件研究所机构知识库
Advanced  
ISCAS OpenIR  > 中科院软件所  > 中科院软件所
题名:
数据空间化的决策支持研究-模型、方法及应用
作者: 石云
答辩日期: 2000
专业: 计算机软件与理论
授予单位: 中国科学院软件研究所
授予地点: 中国科学院软件研究所
学位: 博士
关键词: 空间 ; 地图 ; 数据仓库 ; 联机分析处理(OLAP) ; 数据采掘
摘要: 当前数据库中75%到80%的信息与地理空间位置有关,因此地理空间信息是信息领域中重要的组成部分,更是其它信息系统的基础。若能将地理信息作为组合与搭载其他信息的框架,将数据间的地理关系以清晰简明的地图和图形表现出来,并通过地理分析功能来解决问题,将会使决策分析处理的功能更为强大。目前地理信息系统(GIS)是人类进行地理分析、获取地理信息和地理知识的重要媒体,经过30多年的发展,GIS已经走出特定专业领域,在全球多种领域的信息服务中得到广泛应用。在取得巨大发展的同时,其缺点也越来越明显,其中分析功能的不足一直是制约GIS使用的主要因素。另外,传统的GIS对空间数据的处理是过程化和分离的,其根本原因在于对空间数据没有统一的数据仓库和管理系统,不能将空间数据作为一个整体来对待。这种处理方法限制了空间数据应用的发展。针对上述背景,本文提出了数据空间化的决策支持框架,利用当前决策分析领域中最新的思想和技术,以地理信息作为主线,构建出以空间数据仓库为基础,以空间联机分析处理(OLAP)和空间数据采掘技术为分析手段的新的一体化的决策支持构架。该构架既利用了将数据空间化所带来的在决策分析方面的特殊优势,又弥补了传统GIS系统在分析功能方面的不足,反映空间信息的本质,体现了信息系统的设计观念从处理驱动转向数据驱动的发展趋势。本文的主要研究成果为:1.对于仅隐含地具有地理属性而在实际中却没有空间指针与之关联的关系型数据的空间化问题,本文提出数据仓库地图化的空间OLAP模型及其运算。该模型在实践中被证明具有较高的效率和适用性,对于空间OLAP分析、以及进一步的空间数据采掘具有重要意义。2.针对空间数据仓库建立中突出的实现效率低的问题,本文提出基于MBR方法的一体化的解决方案。其优点是:①在数据仓库空间化的过程中,启发式地进行地理编码,有助于高效地建立空间数据仓库;②自动进行概念层次的划分,便于空间OLAP操作;③有利于空间视图的实体化,操作简便高效;④便于进行进一步的空间数据采掘。3.针对空间数据的不完整和不确定性,提出了基于Rough Set的三阶段空间数据分类方法。实验表明在训练数据集中缺失某些属性的情况下,采用该分类方法比采用通常的决策树算法所生成的规则具有更高的灵敏度,对于解决包含不完整空间信息的问题是有效的。4.对空间OLAP与空间数据采掘技术的集成问题进行了探讨,提出基于影响域的空间OLAM模型,有助于在大型空间数据仓库中交互式地采掘多层次的知识。5.设计并实现地图化智能决策分析产品Easy Map,在中国人民保险公司深圳分公司经理查询及风险分析系统的应用中取得了良好的效果。
英文摘要: It is reported that 75% to 80% of information in the databases possesses the geographical property. Geographic information is not only an important constituent of the information domain, but also the base of constructing other information systems. As a framework to combine and join other information, it can figure out the geographical relation among data via concise maps and graphs, and make the decision analytical functions more powerful. As an important tool for geographical analysis, geographic information system (GIS) has been developed for over 30 years, and has been applied extensively in many kinds of the information service field all over the world. At present, there exist two primary problems in GIS, which restrict its further development. Firstly, its treatment for spatial data is separate and process-oriented, since there is no uniform data warehouse systems for data management. Secondly, in the practical application, it is short of the pertinent analytical method to deal with a large amount of data. In order to solve the aforementioned problems, a spatially enabled decision support framework is put forward. The main thread of the framework is the geographic information hidden in data. It is based on the technology of spatial data warehouse, whose analytic method is the technologies of spatial on-line analytical processing (OLAP) and spatial data mining. The solution not only takes good advantage of the special analytical superiority for spatially enabling the data, but also makes up the defect of the current GIS. It can figure out the nature of the spatial information, and reflect the tendency of the design conception of information system from process-driven to data-driven. The main contributions of the dissertation can be summarized as follows: 1. As for the problem of spatially enabling relational data that just possess the hidden geographical property, a spatial OLAP model for map-enabled relational data warehouse is presented. The operation and implementation of which are also discussed. It has been verified in practice that the model takes on the high effectiveness and flexible applicability. It plays an important role in the field of the spatial OLAP analysis and the further spatial data mining. 2. As for the inefficiency of implementation during the course of constructing the spatial data warehouse, an integrated solution based on the MBR method is presented. It has several advantages. ①The geocoding process can be carried out in an interactive and heuristic way. As a result, the spatial data warehouse can be constructed more effectively. ②It can separate the concept hierarchy automatically, which facilitates the spatial OLAP operation. ③It is in favor of the materialization of the spatial view, which will make the query operation more conveniently and effectively. ④It is favorable to the spatial data mining more in-depth. 3. An effective three-step method, which is based on the rough set theory for spatial data classification, is proposed. The pertinent experimental result shows that the sensitivity produced by the method is higher than that by means of the ID3 algorithm in the case that the training datum runs short of some properties. The validation of this algorithm is verified by the problem of incomplete spatial information. 4. The integrated problem of the spatial OLAP and the spatial data mining technology is discussed, and the spatial OLAM model, which based on the influence domain, is presented. It is favorable to mine multi-level knowledge in an interactive way in a large spatial data warehouse. 5. In practice, a spatially enabled intelligent decision-making tool (Easy Map) is designed and implemented according to the content of this dissertation. It has been running into operation in Shenzhen sub-company affiliated with People's Insurance Company of China and has gained good effects in the system of the manager query and the risk analysis.
语种: 中文
内容类型: 学位论文
URI标识: http://ir.iscas.ac.cn/handle/311060/6798
Appears in Collections:中科院软件所

Files in This Item:
File Name/ File Size Content Type Version Access License
LW002979.pdf(1856KB)----限制开放-- 联系获取全文

Recommended Citation:
石云. 数据空间化的决策支持研究-模型、方法及应用[D]. 中国科学院软件研究所. 中国科学院软件研究所. 2000-01-01.
Service
Recommend this item
Sava as my favorate item
Show this item's statistics
Export Endnote File
Google Scholar
Similar articles in Google Scholar
[石云]'s Articles
CSDL cross search
Similar articles in CSDL Cross Search
[石云]‘s Articles
Related Copyright Policies
Null
Social Bookmarking
Add to CiteULike Add to Connotea Add to Del.icio.us Add to Digg Add to Reddit
所有评论 (0)
暂无评论
 
评注功能仅针对注册用户开放,请您登录
您对该条目有什么异议,请填写以下表单,管理员会尽快联系您。
内 容:
Email:  *
单位:
验证码:   刷新
您在IR的使用过程中有什么好的想法或者建议可以反馈给我们。
标 题:
 *
内 容:
Email:  *
验证码:   刷新

Items in IR are protected by copyright, with all rights reserved, unless otherwise indicated.

 

 

Valid XHTML 1.0!
Copyright © 2007-2017  中国科学院软件研究所 - Feedback
Powered by CSpace